Elektrosurgery bipolar forceps

1. BIOIMPEDANCE OF SOFT TISSUE UNDER
COMPRESSION AND APPLICATIONS TO
ELECTROSURGERY
by
Robert Edward Dodde
A dissertation submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
(Biomedical Engineering)
in The University of Michigan
2011
Doctoral Committee:
Professor Albert J. Shih, Co-Chair
Associate Professor Joseph L. Bull, Co-Chair
Professor James D. Geiger
Assistant Professor Eric Johnsen
Research Investigator Grant H. Kruger

3. © 2011 Robert E. Dodde

4. ii
Dedication
This dissertation is dedicated to my family and our future together; to my beautiful
wife Erica who has shown an incredible amount of strength and faith in myself and us,
has sacrificed incredibly to see me through this, and whom I so longingly look forward to
getting reacquainted with once this is completed, and to my children, Caden and Lawson,
who have given me reasons to laugh when I wanted to cry and made me appreciate the
simple things in life when it has seemed so complex.
I also dedicate this work to those who will not be able to celebrate the conclusion of
this journey with me, but who have meant so much to me during it; my grandfather
Egbert Dodde whose skill at machining inspired me to fabricate much of the hardware for
my research, and my fellow Ph.D. candidate who will always be a philosophical
doctorate to me, Alan Vincent.

5. iii
Acknowledgements
Research of this length and depth cannot happen without the help, guidance, and
support of many people. It is with great appreciation and humility that I offer this work
up for review.
I would like to thank first my advisor, Professor Albert Shih, for his willingness to
take on a graduate student with no engineering background and no financial support. It
was a leap of faith on both of our parts when we began working with each other as
neither of us knew what would happen next. He has given me much freedom to pursue
this research where it needed to go, even when he was not sure that the path I was on
would bear fruit. With his help and guidance, I have not only been able to develop
electrosurgical instrumentation designs, but have received a patent for this research. I
have gained extensive experience writing grant proposals, and have even seen a few get
accepted, which will undoubtedly serve me well as I pursue my future career.
I would also like to also thank my other committee members, Professor Joseph Bull,
Dr. James Geiger, M.D., Dr. Grant Kruger, Ph.D., and Professor Eric Johnsen for their
guidance and critiquing of this research as it progressed. I appreciate their support and
advice throughout this entire process.
Professor Joseph Bull has been incredibly supportive of me as a student and future
researcher. He has always been there to listen and offer advice regardless of the topic of
discussion. It is difficult to feel supported working in a traditional manufacturing
research lab performing biomedical research. Prof. Bull and his lab have adopted me as
one of their own which has been greatly appreciated. With him I have also completed
research not included in this dissertation on the splitting of bubble through a bifurcating
network of channels. To a student such as myself not having an engineering background,
this additional experience is invaluable to me in my future career pursuits and has given
me confidence in my ability as an engineer.

6. iv
Dr. James Geiger, M.D., has been an incredible advocate not only for my research but
for me as an individual. His willingness to work with engineering students on a student
project to measure tissue temperature during electrosurgery is what allowed Prof. Shih
and I to begin working together. He has been an invaluable resource in relating my
research to clinical needs. He has also been a great resource for hardware needed to
perform my experiments. Much of the work presented here would not have been
completed without him.
Dr. Grant Kruger, Ph.D., has proven to be a great blessing to my research. At a point
when the circuitry seemed to be incomprehensible to me and I was near the end of my
patience with this research, Dr. Kruger was there to offer the needed encouragement and
insight I needed to complete the bioimpedance measurement circuitry. I am also grateful
for the additional work I have been able to perform in developing other circuits and
printed circuit boards for ultrasound and pain testing applications. This broadened scope
of electrical work has given me the confidence in my abilities to make and produce my
own circuits.
A special thanks goes out to Professor Eric Johnsen, who made himself available to
serve as my committee cognate on short notice. His willingness to sit and listen to me go
over my defense is much appreciated. Additionally, his critical insights into how the
tissue compression research may be applicable to other areas of much different time and
length scales has opened my mind to the potential impact that this research may have in
the future.
This thesis would also not have been possible without the help of many people from
across the University. First, I would like to give my deepest thanks to Toby Donajkowski.
His willingness to educate and work with me on machining was instrumental in allowing
me to fabricate the components necessary for my research. He never let me settle for a
second rate job on machining a part, even if it was only a block of aluminum to space
parts with. Toby has made me very proud of the work I have done and continues to be a
great example to me of what it means to be a man, husband, and father.
Two University physicians, Dr. William Roberts and Dr. Arnold Advincula, have also
been invaluable to my work. From access to tissue, help in animal labs, and

7. v
conversations about surgical needs, Will is a greatly appreciated resource. Arnie, despite
his busy schedule, has always been such a great cheerleader and together we have done
some great work, including much of the research presented in Chapters 2 and 4. He has
given me access to venture capitalists interested in my research and, despite having left
the University to pursue his career in Florida, continues to support my work.
Professor Kevin Pipe served on both my Master‟s Thesis and Preliminary Exam
committees and unfortunately was not available to participate in the Dissertation Defense.
However, he has always been available with his critical insight into my research at
multiple points along the way. With his keen engineering and analytical mind,
knowledge of both thermal and electrical processes, and desire to see my research be all
that it can be, he routinely could be counted on to give critical insights to the problems I
ran into with my research and the analysis of the data I collected.
Professor William (Rick) Weitzel has been an incredible support and has offered
much insight into areas bioimpedance may have clinical relevance that I had not even
thought of prior to meeting him. We have had a great symbiotic relationship where I
have been able to grow and strengthen my interest in electronics and circuits as well as
become acquainted with ultrasound technologies. His innovative mindset is refreshing
and inspiring.
Scott Merz has also been a great example to me of what it means to be a biomedical
engineer. His intellect and insight into the clinical relevance of experimental research has
been instrumental in how I think about the future potential of my research. We have
worked on a few proposals attempting to elevate my electrosurgical and bioimpedance
research into the clinical space and I have enjoyed the education I have received in
reshaping how I think and present my research to different groups of people.
I would like to thank Gail Rising, Sandy Holmes, and Kimberly Ives for their help in
running animal labs and getting me access to tissue, especially at those times when I
desperately needed it, they were there to support me with their skills and time.
I must give thanks to my fellow lab mates, and a special thanks to Matthew
Chastagner, my ally and friend for many of the years of my research. His decision to join
me in looking at the effects of tissue compression gave me much needed support in

8. vi
making my way through this research. When things were the worst, we were there for
each other to give support, advice, and a listening ear. I would like to thank Jacob Gee,
Roland Chen, and Scott Miller for their help with conducting the experiments within this
thesis. Lastly, I would like to thank Carl McGill and Jason Moore for their support and
encouragement.
Finally, I owe my largest thank you to my family, for their encouragement and
patience throughout this process. It has not been easy and I know it would not have been
possible without my loving wife. I entered into this research married to a incredible
woman and find myself, as my Ph.D. research wrapped up, with two adorable children
who mean everything to me. Erica has sacrificed her career as a social worker to raise
our children and support our family, and for that I am endlessly appreciative. Thank you
from the bottom of my heart for your encouragement and love; I will love you always. To
my parents, who have always been so proud of me, thank you not only for the
encouragement, but for everything you have given me over the years. I owe much of
what I have become to you. To the rest of my family, some of whom I have not even seen
in many years, I love you all and look forward to seeing you again soon.

9. vii
Table of Contents
Dedication ............................................................................................................... ii
Acknowledgements .................................................................................................... iii
List of Figures...............................................................................................................x
List of Tables ........................................................................................................... xvii
List of Appendices.................................................................................................. xviii
Abstract ............................................................................................................ xix
Chapter 1 Introduction................................................................................................1
1.1 Motivation....................................................................................................1
1.2 Literature Review.........................................................................................3
1.2.1 Review of Electrosurgery and Thermal Spread Monitoring ......................3
1.2.2 Review of Electrosurgical Modeling..........................................................6
1.2.3 Review of Thermal Management in Electrosurgical Applications ............6
1.2.4 Review of Tissue Electrical Characterization ............................................7
1.2.5 Review of Tissue Mechanical Property Modeling...................................11
1.1. Research Objectives and Tasks..................................................................12
1.2. Outline........................................................................................................13
Chapter 2 Thermal Profiling of the Energy-Based Surgical Procedure...............15
2.1 Introduction................................................................................................15
2.2 Materials and Methods...............................................................................16
2.3 Results........................................................................................................20
2.4 Discussion..................................................................................................21
2.5 Conclusion .................................................................................................22
Chapter 3 Finite Element Model of the Bipolar Electrosurgical Procedure ........24
3.1 Introduction................................................................................................24
3.2 Experimental Setup for In-Vivo Electrosurgical Temperature
Measurement..............................................................................................27
3.3 Finite Element Modeling ...........................................................................29
3.3.1 Thermal-Electric FEM Formulation.........................................................29

10. viii
3.3.2 Properties of Biological Tissue ................................................................31
3.3.3 FEM Techniques ......................................................................................32
3.3.4 Electrode Design ......................................................................................34
3.3.5 Boundary Conditions................................................................................35
3.3.6 FEM Electrical Input................................................................................37
3.4 Experimental and FEM Results .................................................................37
3.4.1 Experimental Validation and Effect of Compression...............................37
3.4.2 Effect of Temperature-Dependent Electrical and Thermal
Conductivities...........................................................................................40
3.5 Discussion of FEM Results........................................................................41
3.5.1 Temporal and Spatial Temperature Distributions ....................................41
3.5.2 Effect of Compression on Temporal and Spatial Temperature
Distributions .............................................................................................42
3.6 Conclusions................................................................................................44
Chapter 4 Bipolar Electrosurgery Using Active Cooling Channels for the
Minimization of Thermal Spread...........................................................47
4.1 Introduction................................................................................................47
4.2 Materials and Methods...............................................................................48
4.2.1 Spleen Coagulation...................................................................................51
4.2.2 Mesenteric Vessel Sealing........................................................................52
4.3 Results........................................................................................................52
4.4 Discussion..................................................................................................57
4.5 Conclusions................................................................................................59
Chapter 5 Development and Validation of a Bioimpedance Measurement
System Using a New Constant Current Source.....................................60
5.1 Introduction................................................................................................60
5.2 Equivalent Circuit Analog for Soft Tissue.................................................61
5.2.1 Complex and Polar Impedance Representations......................................62
5.2.2 Depressed loci ..........................................................................................63
5.3 Bioimpedance Measurement Circuit..........................................................64
5.3.1 Voltage-Controlled Current Source Circuitry ..........................................67
5.3.2 Current-to-Voltage Converter Circuitry...................................................68
5.3.3 Buffered Differential Voltage Measurement Circuitry ............................69
5.3.4 PCB Layout ..............................................................................................71
5.4 Bioimpedance Measurement Circuit Validation........................................72
5.4.1 Current-to-Voltage Circuit Characterization............................................72
5.4.2 Voltage-Controlled Current Source Characterization ..............................73
5.4.3 Buffered Differential Voltage Measurement Characterization ................74
5.5 Bioimpedance Probe..................................................................................77
5.6 Bioimpedance Probe Characterization.......................................................78
5.7 Hook Effect................................................................................................81

11. ix
5.8 Experimental Methodology .......................................................................82
5.9 Results........................................................................................................84
5.10 Discussion..................................................................................................86
5.11 Conclusions................................................................................................87
Chapter 6 Bioimpedance of Porcine Spleen under Compression..........................88
6.1 Introduction................................................................................................88
6.2 Experimental Set-up and Procedure...........................................................89
6.2.1 Characterization of the Tissue Chamber ..................................................90
6.2.2 Experimental Procedure ...........................................................................93
6.2.3 Experimental Calibration .........................................................................94
6.2.4 Non-linear Least Squares Fitting..............................................................96
6.3 Results........................................................................................................97
6.3.1 Bioimpedance Measurements...................................................................98
6.3.2 Pressure Measurements ..........................................................................100
6.3.3 Histology ................................................................................................100
6.4 Discussions ..............................................................................................103
6.5 Conclusions..............................................................................................105
Chapter 7 Conclusions.............................................................................................107
7.1 Major Achievements................................................................................107
7.2 Original Contributions .............................................................................110
7.3 Future Work.............................................................................................111
Appendices ............................................................................................................115
References ............................................................................................................136

12. x
List of Figures
Figure 1.1. Representation of tissue under compression during electrosurgery. ..............3
Figure 1.2. Frequency-dependance of polarization resistance (RP) and capacitance
(CP) for a 1.4 mm2
platinum electrode.....................................................................8
Figure 1.3. Summary of complex plane and frequency for bioimpedance. ....................10
Figure 2.1. (a) Photograph of exteriorized porcine spleen and (b) in situ
temperature acquisition set-up...............................................................................15
Figure 2.2. Photograph of top and front views of actual polycarbonate fixtures
used for positioning of both the instrument jaws and the thermistors. ..................16
Figure 2.3. Computer Aided Design (CAD) drawing of polycarbonate fixtures
displaying distances of thermistors from instrument jaw edge..............................17
Figure 2.4. Schematic diagram of the temperature acquisition system used. The
signal conditioner consists of a Wheatstone bridge powered by 10V, with
R1=R2=R3=10kΩ..................................................................................................19
Figure 2.5. (a) Post-operative spleen from 5 mm Gyrus Plasmakinetic Cutting
Forceps and (b) post-operative spleen from Ethicon Harmonic Ace.....................20
Figure 2.6. Thermal Profiles for each of the energized instrument trials performed.
(a) Gyrus 5mm Cutting Forceps, (b) Ethicon Harmonic Ace – „Min 3‟
setting for 5 s, (c) Ethicon Harmonic Ace – „Min 3‟ setting until tissue
transaction, and (d) Ethicon Harmonic Ace – „Min 3‟ setting for 5 s
followed by „Max 5‟ setting until tissue transaction..............................................21
Figure 3.1. The Gyrus ACMI 5 mm bipolar cutting forceps. Note the end of the
device is magnified to show electrode detail.........................................................26
Figure 3.2. Experimental set-up showing positioning of tissue, electrode, and
thermistors..............................................................................................................27
Figure 3.3. Voltage input for FEM (a) measured alternating current (AC) voltage
signal and close up view of the 350 kHz waveform and (b) resultant direct
current (DC) approximation of the waveform equivalent to the root-mean-
square (RMS) value of the RF signal.....................................................................29
Figure 3.4. Schematic of the 3D FEM model showing (a) the tissue, electrodes,
and symmetry plane defined by points EACG , (b) a representative mesh
case, (c) top view of tissue regions identified for the compression-dependent
regions (I, II and III) and thermistor......................................................................32

13. xi
Figure 3.5. Two electrodes: (a) Flat electrode (FE), (b) Grooved electrode (GE) in
Gyrus ACMI bipolar instrument used in the experiment, and (c) dimensions
of the cross-section for GE. ...................................................................................34
Figure 3.6. FEM of the effect of temperature-dependent σ and k on tissue
temperature. kTref and k(T) are constant and temperature-dependent thermal
conductivity, respectively. σTref and σ(T) are constant and temperature-
dependent electrical conductivity, respectively. ....................................................36
Figure 3.7. Comparison of thermal profiles for in-vivo experiments and FEM
using a GE under a constant thermal conductivity, temperature-dependent
electrical conductivity, and (a) compression-independent and (b)
compression-dependent simulation........................................................................38
Figure 3.8. Cross-sectional view of temperature profiles on a plane offset from
plane ABDC by 6 mm at different times for a constant thermal conductivity,
temperature-dependent electrical conductivity, and compression-
independent simulation using a GE. Times (a)-(e) correlate to the end of
each pulse and (f) correlates to the end of the simulation after sufficient
cooling....................................................................................................................39
Figure 3.9. Cross-sectional view of temperature profiles at 3.22 s on different
planes offset from plane ABFE for a constant thermal conductivity,
temperature-dependent electrical conductivity, and compression-
independent simulation using a GE (distances indicating the offset from
Plane ABFE) (same temperature scale as in Figure 3.8). ......................................40
Figure 3.10. Cross-sectional view of temperature profiles for a constant thermal
conductivity, temperature- and compression-dependent electrical
conductivity simulation using a GE on a plane offset from plane ABDC by 6
mm at various times. Times (a)-(e) correlate to the end of each pulse and (f)
correlates to the end of the simulation after sufficient cooling..............................41
Figure 3.11. Cross-sectional view of temperature profiles at 3.22 s for a constant
thermal conductivity, temperature- and compression-dependent electrical
simulation using a GE on different planes offset from plane ABFE
(distances indicating the offset from Plane ABFE) (same temperature scale
as in Figure 3.10). ..................................................................................................42
Figure 3.12. Effect of grooved electrode (GE) and flat electrode (FE) on
temperature profiles. ..............................................................................................44
Figure 4.1. Images of the electrodes for the (a) Gyrus 5mm Forceps (standard
bipolar forceps) with thermistor fixture and (b) the modified Gyrus 5mm
Forceps (actively cooled bipolar forceps) incorporating an active cooling
channel around the outside of the bipolar electrodes with thermistor fixture
attached. .................................................................................................................48
Figure 4.2. Schematics for the two devices used in this experiment. (a) Standard
bipolar forceps and a top view showing thermistor fixture and position of
the thermistors. (b) Actively cooled bipolar forceps incorporating cooling

14. xii
channels around both bipolar electrodes and top view showing thermistor
fixture and position of the thermistors...................................................................49
Figure 4.3. Schematic diagram of data acquisition system for electrical and
temperature data recording. The experimental components, including the
electrosurgical system and tissue, are shown within the dashed rectangle.
Thermistors were inserted into the tissue to record temperature
measurements while the current and voltage probes were connected to the
leads of the electrosurgical devices to record electrical measurements. Data
was brought into the computer via a NI PXI-6221 Data Acquisition Card
and an Agilent 54833A Oscilloscope. ...................................................................50
Figure 4.4. Experimental set-up for coagulation experiments using (a) the
unmodified Gyrus 5mm Forceps and (b) the modified Forceps incorporating
an active cooling channel around the bipolar electrodes. ......................................52
Figure 4.5. Typical thermal and electrical profiles for coagulation studies on
splenic tissue using the standard bipolar forceps and the actively cooled
forceps with a cooling channel around the bipolar electrodes. (a) Standard
bipolar forceps. Midline temperature reached 90% of maximum value in 2.5
s. Average power for first 2 pulses was 125 W and dropped down to 40 W
by the 5th and 6th pulses (indicative of the VP3 40 setting). Impedance
averaged 450 Ω over the 4th-6th pulses. (b) Actively cooled forceps.
Midline temperature reached 90% of maximum value in 2.5 s. Average
power for first 4 pulses was 140 W and dropped down to 85 W for the 6th-
8th pulses. Impedance remained lower than 50 Ω for the first 4 pulses and
averaged 200 Ω over the 6th-8th pulses.................................................................54
Figure 4.6. Typical thermal and electrical profiles for coagulation studies on
mesenteric tissue using the standard bipolar forceps and the actively cooled
forceps with a cooling channel surrounding the bipolar electrodes. (a)
Standard bipolar forceps. Midline temperature reached 90% of maximum
value in 4 s. Average power for first 3 pulses was 135 W and dropped down
to 30 W by the 7th and 8th pulses (indicative of the VP1 30 setting).
Impedance slowly rose to ~240 Ω in the later pulses. The steeper
temperature drop-off at the 1.0 mm point once power input stopped
indicates maximum temperatures were reached not in the midline but
outside the profile of the forceps. Temperature data not available for the 3.5
mm distance. (b) Actively cooled forceps. Midline temperature reached
90% of maximum value in 7 s. Average power for first 8 pulses was 110 W
while impedance never reached more than 40 Ω...................................................55
Figure 4.7. Summary of thermal and temporal development for coagulation
experiments on splenic (a) and mesenteric (b) tissues. (a) SPLN data comes
from experiments using standard bipolar forceps while the SPLC data comes
from experiments with the actively cooled forceps. Midline temperatures
were seen to increase by a statistically significant larger amount (p-value =
0.009) with the actively cooled forceps as compared to the standard bipolar
forceps while temperatures taken adjacent to the actively cooled forceps

15. xiii
increased less than 2°C. Adjacent points to the standard bipolar forceps saw
temperature changes slowly decline from 30°C at 1.0 mm to 4°C at 3.5 mm.
(b) MESN data comes from experiments using standard bipolar forceps
while the MESC data comes from experiments with the actively cooled
forceps. Midline temperatures were seen to increase slightly more in the
cooled case as compared to the standard case while temperatures taken
adjacent to the cooled device rose minimally. Adjacent points to the
standard bipolar forceps saw temperature changes slowly decline from 40°C
at 1.0 mm to 10°C at 3.0 mm (data not available for 3.5 mm distance)................56
Figure 4.8. Detailed summary of thermal development during each pulse for
coagulation experiments on splenic (a) and mesenteric (b) tissues. (a) SPLN
data comes from experiments using normal 5 mm forceps while the SPLC
data comes from experiments with the custom adjacent cooling channels.
Midline temperatures were seen to increase significantly more during each
pulse in the cooled case as compared to the standard case while
temperatures taken adjacent to the cooled device rose minimally. Adjacent
points to the normal device saw temperature changes increase even more at
the 1.0 mm location compared to the midline temperatures during the first
pulse. More distant locations saw similar slowly declining slopes for
temperature increases during each pulse through 1.5 mm. Points at 3.0 mm
and 3.5 mm saw greater temperature increases at later pulses due to
conductive heating. (b) MESN data comes from experiments using normal
5 mm forceps while the MESC data comes from experiments with the
custom adjacent cooling channels. Midline temperatures were seen to
increase at nearly equal rates, although the standard device heated up
quicker during the beginning pulses and the modified device heated up more
quickly during the later pulses. However, hardly any temperature increase
was seen at points adjacent to the cooled device while similar, shifted
temperature profiles were seen for the standard device at distances adjacent
to the tool edge.......................................................................................................58
Figure 5.1. Overview of how alternating current passes through cellular tissue. ...........60
Figure 5.2. Schematic of tissue as an equivalent circuit analog. Rext is the
resistance of the extracellular fluid, Rint is the resistance of the intracellular
fluid, Cm is the bulk capacitance of the cellular membranes, and Rm is the
bulk resistance of the cellular membranes.............................................................61
Figure 5.3. (a) Overview of polar and complex representations of data (b)
equivalent circuit diagram for tissue......................................................................62
Figure 5.4. Example of depressed loci in the complex plane. R represents
resistance and X represents reactance....................................................................63
Figure 5.5. Schematic overview of bioimpedance circuitry. Circuit components
are located within the dashed box. I_OUT, V1, V2, and I_IN connect to the
probe and then to the tissue. VS is the signal source from the function
generator. V(V) is the voltage signal proportional to the voltage difference
between V1 and V2. V(I) is a voltage signal proportional to the current

16. xiv
injected into the tissue. V(V) and V(I) connect to an oscilloscope for
measurement. .........................................................................................................66
Figure 5.6. Circuit schematic for (a) the voltage-controlled current source (VCCS)
used to drive the tissue load and (b) the current-to-voltage converter (C2V)
used to monitor the input current from the VCCS.................................................69
Figure 5.7. Detailed schematic of the buffered voltage differentiator circuit. ................70
Figure 5.8. (a) Top side view of the bioimpedance measurement PCB. And (b)
bioimpedance measurement PCB encased in housing unit....................................71
Figure 5.9. Characterization of the C2V subcircuit for the bioimpedance
measurement circuit. (a) Measured current for various input voltages and (b)
deviation of phase shift from expected value of 180°. The black lines in (b)
represent one degree of phase shift........................................................................72
Figure 5.10. Characterization of the VCCS subcircuit for various targeted output
currents...................................................................................................................74
Figure 5.11. Bandwidth details for the differential voltage amplifier circuit with a
gain of 20 in (a) dB and (b) phase shift. ................................................................75
Figure 5.12. Cole-Cole plots for R=1 kΩ (red), R=2 kΩ (blue). (a) C = 100 pF
(102.2 pF measured) (b) C = 1 nF (1.04 nF measured) (c) C = 45 nF (44.4
nF measured) (d) C = 100 nF (98 nF measured)....................................................76
Figure 5.13. (a) Bioimpedance probe and (b) exploded view of the probe. .....................77
Figure 5.14. Estimated contact impedances based on Schwan data [39] assuming a
0.05 mm2
platinum electrode. Total contact impedance for the circuit would
be double................................................................................................................78
Figure 5.15. Experimental (|Z|exp) vs theoretical (|Z|theo) impedance values for NaCl
solutions.................................................................................................................80
Figure 5.16. Experimental Set-up (a) and process flow diagram and (b) for
bioimpedance measurements. ................................................................................83
Figure 5.17. (a) Raw (raw), corrected (IA and IACp), and fitted data (Zfit) for one
example bioimpedance measurement on porcine spleen tissue and (b)
Comparison of electrical conductivity (σ) and electrical permittivity (ε)
experimental results for porcine spleen with the literature (note literature
based on a range of animal spleen values at various temperature ranges).............85
Figure 6.1. Experimental set-up for tissue compression experiments.............................89
Figure 6.2. (a) Measured |Z| versus frequency for various depths above the bottom
of the tissue chamber and (b) Average current at each frequency point for
given saline concentrations....................................................................................90
Figure 6.3. Effect of insertion depth (h) on |Z|. (a) At higher frequencies, effect is
seen to be concentration dependent, but for frequencies below 100kHz, the
effect is constant. (b) Average normalized impedance vs. insertion depth for

17. xv
0.01, 0.02, 0.05, and 0.125% saline concentrations (frequency range of
readings is from 100Hz – 100kHz)........................................................................91
Figure 6.4. Impact of probe insertion depth (h) on the measured phase angle. (a)
For high frequencies, impact is concentration dependent. (b) For the
frequency range 100 Hz – 100 kHz, the average phase shift is seen to be less
than one degree regardless of insertion depth........................................................92
Figure 6.5. Dielectric properties for saline solutions versus frequency. (a)
electrical permittivity (ε) and (b) electrical conductivity (σ).................................93
Figure 6.6. Decision and process flowchart for compression of spleen tissue................94
Figure 6.7. Progression of data from the initial collection to the final fitting to the
Cole-Cole model. Raw data (Raw), data after calibrating for voltage
differentiator (IA), data after further correcting for depth effect (IAd), after
next correcting for the Hook Effect (IAdCp), and the final fitted data (Zfit).
Data was collected at a height (h) of 6.20 mm. .....................................................96
Figure 6.8. Final summary of strain-dependent Cole-Cole terms. (a) Normalized
Rext and Rint as a function of compressive ratio, (b) normalized Cmem as a
function of compressive ratio, and (c) normalized α as a function of
compressive ratio. ..................................................................................................97
Figure 6.9. Mean and standard deviation (error bars) of the RMSE for data fitting
to a Cole-Cole model for eight samples over 0-80% compression........................98
Figure 6.10. Overview of the development of the impedance loci during tissue
compression. Note the shift of the loci to the right for lower compression
levels and then a dramatic shift to the left and collapse of the loci at higher
compression levels.................................................................................................99
Figure 6.11. Plot of pressure (σ) versus time (s) for strain measurements. Each
ramp up of pressure and subsequent relaxation is for an incremental increase
in strain of 10%. X-axis is in time (s), with the length of the last relaxation
for 80% compression being 120 s........................................................................100
Figure 6.12. Histology at 10X magnification for (a) 80% tissue compression and (b)
control. .................................................................................................................101
Figure 6.13. Histology at 40X magnification for (a) 80% tissue compression and (b)
control. .................................................................................................................102
Figure 6.14. Plot showing the strain dependence of the resistance at 500 kHz.
Compression is defined as where hm is the measured tissue
thickness and hi is the initial tissue thickness. This is a common frequency
used in electrosurgical instrumentation where large strains are commonly
imposed on tissue. From 50-80% compression there is seen to be a 31%
change in the resistance of the tissue...................................................................104
Figure A.1. Load curves for each of the VPC modes within the Gyrus
PlasmaKinetic®
Generator. ..................................................................................117

18. xvi
Figure B.1. (a) Overview of polar and complex representations of data and (b)
equivalent circuit analog for tissue. .....................................................................118
Figure B.2. Graphical view of transformations required to represent complex data
in the impedance, Z*, admittance, Y*, modulus, M*, and capacitance, C*,
planes. ..................................................................................................................119
Figure C.1. Block diagram of the Labview program. The numbered rounded
rectangles indicate (1) while loop, (2) case structure, (3) shift register, (4)
cluster, (5) array, and (6) enumerated text box....................................................125
Figure C.2. Detailed view of (a) the enumerated text box showing all of the states
within the program and (b) the cluster containing all of the different data
type variables used in the program. .....................................................................126
Figure C.3. Front panel of the bioimpedance Labview program...................................127
Figure D.1. Full schematic for bioimpedance measurement circuit. DC bypass
coupling, BNC connections, and offset voltage potentiometers are shown
along with power supply lines for clarification. ..................................................129
Figure D.2. (a) Eagle board layout of bioimpedance measurement circuit (b)
bottom view of final PCB board (c) top view of final PCB board. .....................130
Figure D.3. AMP03 datasheet........................................................................................131
Figure D.4. AD711 datasheet.........................................................................................132
Figure D.5. AD8065 datasheet.......................................................................................133
Figure D.6. INA111 datasheet........................................................................................134
Figure D.7. LM7171 datasheet.......................................................................................135

19. xvii
List of Tables
Table 2.1. Comparison of Two Energized Surgical Instruments.......................................18
Table 2.2. Comparison of mean peak temperatures, standard deviations, and
procedural times for the 5 mm Gyrus Cutting Forceps and Ethicon
Harmonic Ace........................................................................................................22
Table 3.1. Properties used in the FEM...............................................................................31
Table 3.2. FEM boundary conditions as marked in Figure 3.4..........................................35
Table 5.1. Summary of bioimpedance probe calibration data ..........................................81
Table A.1. Summary of VPC modes within the Gyrus PlasmaKinetic®
Generator. .......116

20. xviii
List of Appendices
Appendix A Gyrus PlasmaKinetic®
Surgical Generator Overview.............................. 116
Appendix B Cartesian and Polar Bioimpedance Representations ................................ 118
Appendix C Bioimpedance Labview Program............................................................. 122
Appendix D Bioimpedance Measurement Circuit Details............................................ 128

21. xix
Abstract
This research studies the impact of compression on the electrical impedance of soft
tissue. Soft tissue compression occurs in many ways, either physiologically or through
forces imposed by the outside world. While much work has been done in characterizing
the electrical properties of soft tissues, little work has been done in correlating the
compression of tissue to the subsequent changes in the electrical properties. This lack of
knowledge has lead to the use of less than ideal instrumentation within electrosurgery
resulting in unnecessary quality of life reduction in patients post-operatively.
This research aims to quantify the impact of compression on the bioimpedance of
tissue. First, a baseline for understanding thermal spread is developed by documenting
the thermal history of tissue during energized surgical procedures. Then, a finite element
model of the bipolar electrosurgical procedure is performed and analyzed to suggest
various mechanisms that can be at play during this procedure that affect the resulting
thermal profile. Next, a new instrument design is developed and tested based on these
findings to eliminate thermal spread. To perform more detailed research on electrical
impedance changes in compressed tissue, a bioimpedance measurement system is then
developed and validated. Finally, basic tissue compression tests are performed where
mechanical and electrical properties of the tissue are monitored during the tissue
compression. These findings are finally correlated to suggest a mechanism for the
process of tissue compression.

22. 1
Chapter 1 Introduction
1.1 Motivation
Over the last two decades dramatic advances in surgery have allowed increasingly
complex operations to be completed using minimally invasive surgery (MIS). MIS is
accomplished by passing laparoscopic ports through very small incisions. Specialized
instruments are passed through the ports to complete the operation. A number of
potential advantages are thus offered including reduced pain, scarring, and convalescence
periods. Advancements in energy-based dissection and vessel sealing (ligation) devices
have been critical to broadening the MIS applications. These energy-based surgical
devices (EBSD) use a variety of energy sources (electrical, radiofrequency, ultrasonic,
and laser) to provide effective and rapid hemostasis and are approved for ligating blood
vessels up to 7 mm in diameter [1]. Use of EBSDs has allowed successful dissection of
very vascular organs such as the prostate, liver and uterus. They are now adapted for
nearly all types of surgery including neurosurgery, orthopedics, gynecology, urology,
general surgery, plastic surgery, and otolaryngology. While originally developed for
laparoscopic surgery, EBSDs are now used in open surgery where they are replacing
traditional suture ligation of blood vessels. EBSD success has been tempered somewhat
by recognition of possible collateral tissue damage due to thermal or electrical spread
from the instrument tip.
These EBSDs are used in thousands of operations yearly and complications
associated with them are thus magnified. Annually 600,000 hysterectomies and 220,000
prostatectomies (29,000 deaths) are performed and they are the second most widely
performed surgeries among respective genders [2]. These patients are also younger and
healthier than those diagnosed in the past. Side effects from these procedures include
urinary incontinence (UI) in males and females and erectile dysfunction (ED) in males. A
study performed by Brown et al. [3] suggests the extent of uterine removal, and thus
increased potential for neurovascular bundle (NVB) damage, is correlated to risk of UI. A

23. 2
technique to spare NVBs around the prostate has been developed. Widespread adoption
of this technique (meticulous dissections and preservation of the neurovascular bundles
without use of electrosurgical devices) has resulted in improved post-operative potency
rates of 68% to 86% at centers of excellence [4, 5]. Within the last ten years, advanced
laparoscopic and robotic techniques have been applied to these procedures, including
laparoscopic radical prostatectomy (LRP). Although long-term data are not yet available,
results with respect to biochemical progression free survival and continence appear
equivalent to open techniques. However, a great variability is still seen in postoperative
potency rates [6]. The majority of published techniques of laparoscopic prostatectomy
rely on bipolar, monopolar, or ultrasonic energy sources for hemostasis. Successful
preservation of potency has recently been reported from centers specializing in LRP with
results ranging from 23% to 82%. This wide variability is likely multifactorial but may be
in part due to the method of nerve dissection and use of hemostatic energy sources.
Additionally, it has been demonstrated by Ong et al. in a canine model that dissection
utilizing any EBSD can result in NVB damage and poor post-operative potency outcomes
due to large thermal spread [7].
It has been suggested that one of the primary reasons for these complications is due to
the excessive thermal energy developed in the region during energized surgical
procedures [8]. Considering the high compression levels (>60%) used during these
procedures, complex mechanical changes occur which are hypothesized to also impact
other tissue characteristics such as fluid distribution and the electrical conductivity of the
tissue.
This research aims to quantify the impact of high compression on the electrical
properties of tissue. First, a baseline for understanding thermal spread is developed by
documenting the thermal history of tissue during energized surgical procedures. Then, a
finite element model of the bipolar electrosurgical procedure is performed and analyzed
to suggest various mechanisms that can be at play during this procedure that affect the
resulting thermal profile. Next, a new instrument design is developed and tested based on
these findings to eliminate thermal spread. To perform more detailed research on
electrical impedance changes in compressed tissue, a bioimpedance measurement system
is then developed and validated. Finally, basic tissue compression tests are performed

24. 3
where mechanical and electrical properties of the tissue are monitored during the tissue
compression. These findings are correlated at the end to suggest a mechanism for the
process of tissue compression.
1.2 Literature Review
1.2.1 Review of Electrosurgery and Thermal Spread Monitoring
Collateral thermal damage, or thermal spread, from energized surgical instruments
has been shown to vary widely by the instruments being used, the modality being used,
the environment they are being used in, and the tissue they are being used on.
It has been well documented that tissue damage occurs thermally through an
Arrhenius-type dependence on temperature and time [9]. The longer tissue is held at a
certain temperature above a threshold, the greater amount of damage occurs, and the
amount of time a tissue can withstand a temperature decreases rapidly with increasing
temperature. Industry standards take 43°C as the starting temperature where damage can
occur at, but it takes a considerably long time to cause permanent damage, on the order of
Figure 1.1. Representation of tissue under compression during electrosurgery.

25. 4
200 minutes. In contrast, tissue held at 65°C for as short a time as one second can cause
permanent damage [10].
Electrosurgery is the descendant of electrocautery and differs primarily from its
predecessor in that it passes current into the body to generate heat using the resistance of
the tissue instead of using electric current to directly heat the tip of the electrode [10]. It
is largely considered to be developed by Harvey Cushing and William T. Bovie in 1926
but others have had major contributions, including William Clark who first described the
desiccating properties of electrosurgery in 1914 [11].
In monopolar electrosurgery, the whole patient becomes part of the circuit. Current
from the active electrode seeks the shortest path back to the return electrode. This
concentrates current along less electrically resistive conduits such as nerves or vascular
structures. The path current takes may therefore have no correlation with anatomical
distance. Partly due to this, the bipolar electrode was introduced by Jeffrey Greenwood in
1942 to eliminate the patient from the circuit [12]. The active and return electrodes are in
close proximity, thus minimizing potential injury. However, as tissue is desiccated, its
resistance increases. Current may spread to surrounding lower resistive tissue and widen
the area of damage.
Additionally, current leakage from the bipolar device may cause it to behave like a
monopolar electrical source. Bipolar cautery has been associated with thermal spread up
to 6 mm [13]. Newer forms of bipolar energy devices, using a generator with a feedback-
controlled circuit to determine the conductivity and amount of tissue in the instrument‟s
grasp, deliver appropriate amounts of high-current, low voltage energy to seal tissue. The
high current melts collagen and elastin in the tissue, forming a durable, thin seal. Liga-
Sure™ Lap (LS) sealing device (Valleylab) and the PlasmaKinetics™ (PK) sealer (Gyrus
Medical) are touted to reliably seal vessels up to 7 mm in diameter with minimal thermal
spread. The Ligasure device uses continuous bipolar waveforms while the Gyrus-PK
sealer uses pulsed bipolar waveform, with inactive periods between energy bursts. Gyrus
theorizes this pulse/cool-off period allows instrument jaw cooling, thus reducing
desiccation at the contact point, and resulting in less electrode sticking. While thermal
spread with these devices is less than traditional bipolar it is still significant and can be

26. 5
greater than 3mm. Prior research examining electrosurgery has identified heat as a cause
of neural injury, which may occur as low as 41°C [7]. Because of this, irrigation is
routinely used to cool tissue during head and neck surgery. This practice is corroborated
by studies finding clear neuro-protective effects with using irrigation during bipolar
cautery of the rat sciatic nerve.
Ultrasonic shears have been developed to obviate the need for electrical current and
may be more suitable for use around neural structures. However, in pre-clinical and
clinical studies it is clear that thermal spread may be significant, often as much as 3-4
mm, and can lead to complications including hollow viscous (bowel, bladder, etc)
perforation and significant nerve injuries. For example, many surgeons began to use the
ultrasonic shears for laparoscopic radical prostatectomy due to its excellent haemostatic
properties. However, Owaki et al. found the ultrasonic shear blades become hot,
increasing to 63°C in 3 s and 150°C after 30 s [14]. They suggested blade contact with
neural structures immediately following use caused recurrent injury in their patients
undergoing endoscopic parathyroid surgery. This is important as the surgeon has no
indication of the instrument tip temperatures or the degree of thermal spread during
laparoscopic surgery.
Methods for monitoring the thermal history of tissue have been varied. Currently,
thermographic imaging techniques are being widely used to demonstrate surface
temperatures of both the tissue and the electrode during procedures. Thermocouple and
thermistor use has also been used in the past, but restricted to imbedding the sensor into
the electrodes themselves. Post-operatively, histopathologic studies can also be
performed and used in coincidence with thermal data to correlate damages seen
histologically and thermally.
The extent of the collateral damage done using energized procedures has been studied
previously with varying degrees of cross-correlation [15, 16]. Traditionally the thermal
spread from these instruments is determined with a combination of in-situ dynamic
thermography and histopathologic studies [17].

27. 6
1.2.2 Review of Electrosurgical Modeling
Thermal spread in biological tissue is difficult to measure and predict. Modeling is a
necessary tool to understand temperature distribution and tissue damage from thermal
spread. However, research is lacking in this area. Research in the modeling of
radiofrequency (RF) ablation has been reviewed by Berjano [18]. Past research has
focused on FEM of tissue RF ablation and shortening the design time for new RF
instrumentation. Several researchers have modeled RF ablation using a finite element
approach [19-21]. However, the literature has been limited primarily to the area of tumor
ablation in the liver and heart, which is characterized by low voltage inputs and
procedural times on the order of 480–720 s and 60-120 s, respectively. Cautery is a
technique characterized by high voltage inputs and procedural times on the order of 3–10
s, almost two orders of magnitude shorter than liver RF ablation and one order shorter
than cardiac RF ablation. To date, detailed FEM on the cautery procedures is still new
and not well studied. Pearce et al. [22] published the finite difference determinations for
the potential gradient from a smooth rectangular electrode. The modeling of such
procedures can be important to further the understanding of how tissue responds to RF
energy, resulting in improvements to instrumentation design.
1.2.3 Review of Thermal Management in Electrosurgical Applications
To mitigate the natural consequences of energized surgical devices, various
techniques have been utilized. These include the use of saline irrigation, internal heat
pipes, and control algorithms within the generator.
Saline flushes have found good use for hepatic tissue ablation where deep, wide
lesions are needed [23-27]. Cauterized tissue at the surface of the liver creates an
impeditive barrier to further cauterization and is therefore unwanted. By flooding the
area with saline, both the surface of the tissue and the electrode are maintained at lower
temperatures, keeping the tissue conductive to electrical current longer, thereby
increasing the lesion depth with time [28]. This technique has recently been advanced
through the use of cold irrigation to further prohibit lateral thermal damage to the tissue
[29, 30].

28. 7
Heat pipes have been used with the development of hollow electrodes to promote
passive cooling of the electrodes primarily in response to tissue charring and a resultant
sticking of tissue to the electrode which results in risks of breaking open sealed and/or
cauterized tissue once the procedure has completed [31, 32].
With the onset of solid-state circuitry in the 1970‟s, surgical generator design began
to advance rapidly [10]. Most electrosurgical generators employ a constant power
control on the electrical output to the surgical devices. This serves to automatically
adjust output voltage to the changing electrical properties of the tissue during a
procedure. Additional proprietary algorithms are also incorporated into most modern
surgical generators [33].
Of note is an additional technology being currently used to control thermal spread.
EnSeal uses a micro-carbon electrode along with an I-beam compressor to coagulate
tissue [34]. The electrode is proposed to microscopically short-circuit at high
temperatures, thus avoiding additional electrical current from being injected into the
tissue.
1.2.4 Review of Tissue Electrical Characterization
Electrical stimulation and measurement had been initially limited to frequencies
above 5 kHz for safety reasons expressed by Geddes et al. [35]. At 50 Hz and using trans-
thoracic electrodes, currents of 50 mA stimulate the Vegas nerve producing a slowing of
the heart. At 5 kHz, about 12 to 15 times more current (600 to 750 mA) is needed to
produce the same effect. At 3 kHz, the current required to produce a ventricular
fibrillation (approximately 4-6 A) is 20 times greater than that at 60 Hz (200-300 mA).
With recent advances in instrumentation technology, it is now possible to use currents in
the μA range, thus permitting safe physiological measurements even at low frequencies.
Contact Impedance
The ability to fully characterize the electrical response of tissue and electrolytes is a
relatively recent phenomenon [36]. Up until the turn of the 20th
century, it was difficult
to measure even the impedance of simple electrolytes accurately due to frequency and
current-density dependencies of the contact impedance of the electrode/electrolyte
interface. In 1897 Kohlrausch minimized the electrode impedance problem by

29. 8
developing the platinum-black electrode, which has a low impedance and allowed for
resistivity measurements at 1000 Hz with a bipolar electrode [37]. However, in 1884
Bouty had already resolved the electrode/electrolyte impedance problem in a different
way by introducing the tetrapolar method [38]. In this method, a constant current is
injected between two outer electrodes and a potential difference is measured between two
inner electrodes each having input impedance much greater than the contact impedances
of the potential-measuring electrodes. Bouty called the potential-measuring electrodes
„parasitic electrodes.‟ It is common practice to use the terms polarization resistance and
polarization capacitance to describe the processes involved with current moving from an
electrode to an electrolyte.
Since the discovery and subsequent understanding of contact impedance, much work
has been performed on the interaction between electrodes and electrolytes [39-42]. In
general, the contact impedance can be modeled as an equivalent electrical circuit by
placing a resistor and capacitor in parallel with each other. As mentioned above, the
Figure 1.2. Frequency-dependance of polarization resistance (RP) and capacitance (CP)
for a 1.4 mm2
platinum electrode.

30. 9
terms are called the polarization resistance (RP) and the polarization capacitance (CP).
The total impedance that the electrode/electrolyte interface has to current can be
expressed as the sum of the separate impedances, or
1-1
where ZP is the polarization impedance.
As shown in Figure 1.2 [39], RP and CP, and thus ZP, are frequency dependent.
However, RP and CP are related to each other through the following equation
1-2
which is fairly frequency-independent, as shown in Figure 1.2 [39].
The Cell Membrane and Bioimpedance
It wasn‟t until the early 1930s that it was understood how critical a component the
reactance of tissue is to biological impedance measurements. Early efforts of researchers
such as Philippson [43], Fricke [44], McClendon [45, 46], and Cole [47-49] demonstrated
the presence of a cell membrane capacitance. Using the potential theory originally
developed by Maxwell [50] which was later applied to particle suspensions by Fricke [51,
52], these investigators confirmed the presence of a „relaxation-type‟ conduction process
associated with the cell membrane capacitance.
Equivalent Circuit for Bioimpedance
Just prior to the establishment of the membrane capacitance, Carter published work in
1925 on the representation of two terminal impedance networks containing reactive
elements [53]. In this work Carter represented impedance as a variable in the complex
plane and it was soon quickly adopted by Cole as an analysis method for biological
impedance [54]. Together with equivalent circuit analogues, this representation of
biological impedance has greatly simplified an otherwise difficult task of analyzing
complex electrical interactions. In 1941, Cole and Cole published what has become
known as the Cole-Cole model for biological tissue [55, 56] which follows the form

31. 10
1-3
where Z is the impedance, ω is the natural frequency, R∞ is the resistance at infinite
frequency, R0 is the resistance at zero frequency (DC), j is the , τ is the relaxation
constant for the system, and α is the dispersion coefficient for the system. It should be
noted that there is discrepancy in who originally posted this relationship, with some
giving credit only to Kenneth Cole, citing a 1940 paper he published [57].
Plots of the real (resistance) versus the imaginary (reactive) parts are semicircular in
nature having centers which lie below the real axis. Also, plotting the real and imaginary
parts separately against frequency provides characteristic shapes as well. Figure 1.3
shows a graphical summary of these plots and how they relate to each other. Analyzing
Eq. 1-3 as a function of frequency, the impedance can be shown to equal R0 as ω
approaching zero Hz (DC). This is due to the high cell membrane resistance in relation
to the interstitial fluid, along with the capacitance of the cell membrane acting as an open
circuit as low frequency [58]. As ω approaches infinity, the impedance becomes equal to
Figure 1.3. Summary of complex plane and frequency for bioimpedance.

32. 11
R∞, due to the cell membrane capacitance acting as a short circuit allowing current to
cross the cell membrane unimpeded.
Modern Uses of Bioimpedance
Today bioimpedance measurements are used in a variety of ways and have matured
beyond a research tool. Bioelectrical impedance analysis (BIA) is currently used to
determine total body water and body fat in individuals. Multi-frequency bioelectrical
impedance analysis (MFBIA) uses impedance measurements over a range of low to high
frequencies to determine the extracellular fluid volume (low frequencies) and total body
fluid volume (high frequencies) [59, 60]. This practice is also being used on a partial, or
segmental, portion of the body for the characterization of edema in dialysis [61] and to
determine hydration status and predisposition to hypotension [62].
Bioimpedance tomography, or electrical impedance tomography (EIT), has matured
to the point that it can be used to produce precise 3D images of the body [63]. EIT does
not have the spatial resolution of MRI or CT, but has a key advantage in its temporal
resolution, which can be down to the order of milliseconds [64]. This technology has
been used in areas such as the detection of breast cancer and the monitoring of brain
function and stroke [65, 66]. While it has yet to be used routinely in everyday clinical
practice, it has great value in being both safe and cheap compared to alternative imaging
technologies [63].
1.2.5 Review of Tissue Mechanical Property Modeling
Initial looks at tissue can take an equivalent mechanical model approach by
combining springs and dashpots in series and in parallel with each other. Major problems
with these approaches arise when creep and relaxation are looked at, where each has its
own flaws. Fung‟s Biomechanics is classical literature describing the quasi-linear
viscoelasticity (QLV) of soft tissue [67]. An excellent review was performed by
Humphrey [68] on the continuum biomechanics of soft tissue.
Since the introduction of this theory, much research has been performed analyzing
tissues from this approach, whereby a range of relaxation times are seen to drive the
tissue response to strain. The two main areas that QLV theory has been put to use are in
the modeling of tendons and ligaments [69-71] and in the modeling of abdominal soft

33. 12
tissue [72-75]. Much of this work continues to be performed for use in developing
realistic computer-based surgical software as well as in the design of surgical robots and
their associated tools.
Considering the number of parameters that must be fit to solve for the QLV model (as
many as eight or more depending on the functions that are defined), a number of papers
have been written concerning the solution of the QLV problem [76-78].
Apart from Fung‟s QLV theory, work done by Mow [79] and more recently carried
on by Ateshian [80, 81] has looked at the harder of the soft tissues, primarily cartilage,
from the standpoint of tissue being viscoelastic. A model, termed the biphasic model, has
been developed describing how these tissues respond to stress and strain which shows
marked differences with QLV theory. An important outcome of this research is the
determination of the protective qualities that fluid has in absorbing stress due to high
strains.
The bioimpedance of tissue undergoing compression has natural logical connections
with the mechanical properties of tissue. Tissue compression results in the loss of fluid
and solid material from the compressed space simply by volume reduction. This impacts
cell membranes, interstitial fluid, and intracellular fluid, which are the principal
components involved in the electrical relaxation seen in bioimpedance readings of tissue
[82]. Monitoring both bioimpedance and the relaxation pressure of tissue can lead to
improved understanding of the mechanics underlying tissue compression.
1.1. Research Objectives and Tasks
The primary objective of this research is to characterize the electrical properties of
tissue under compression and correlate these to the mechanical changes seen in tissue
under compression. As discussed above in the motivation (Section 1.1) collateral thermal
damage resulting from electrosurgery is primary concern for using these devices in
critical surgical areas such as the reproductive areas for prostatectomy and hysterectomy.
Secondary objectives accomplished throughout this process include:
a) the development of a subsurface temperature sensing system for monitoring tissue
temperature during energized surgical procedures

34. 13
b) the finite element modeling of a bipolar electrosurgical procedure
c) the development of a surgical thermal management system for electrosurgery
d) the development of a custom experimental procedure for monitoring
bioimpedance under compression
1.2. Outline
This dissertation presents the concept and application of a compression-dependent
electrical conductivity of tissue. Layout of this dissertation is described in the following
paragraphs.
Chapter 1 of the dissertation provides primary motivations for this doctoral research
as well as a literature review of related research work.
Chapter 2 describes a method for measuring the real-time temperature of tissue during
the use of energy-based surgical devices. The ability of capturing sub-surface
temperatures during energized surgical procedures is validated.
Chapter 3 undertakes the finite element modeling of the bipolar electrosurgical
procedure. Previous knowledge on the modeling of monopolar and RF ablation
techniques is expanded here to include the bipolar technique, with improvements in
matching to the experimental results seen by allowing for a theoretical strain-dependent
electrical conductivity.
Chapter 4 is comparative research performed with bipolar electrosurgical instruments
validating a method for combating thermal spread. Active cooling channels are utilized
adjacent to bipolar electrodes to act as a hear sink and an non-energized compressor to
adjacent tissue. The results not only show a dramatic decrease in thermal spread laterally
from the bipolar device, but also indicate a more efficient coagulation process as tissue
temperatures at the center of the actively cooled device get hotter more quickly than for
the standard bipolar device.
Chapter 5 leverages the use of the four electrode technique to analyze the real-time
electrical conductivity of tissue. The idea of compression-dependent electrical
conductivity is introduced. A series of probes are tested, combined with an LCR meter
and a custom front-end amplifier, in ex-vivo tissue to validate the approach. The chapter

35. 14
culminates in both ex-vivo and in-vivo studies analyzing the electrical conductivity of
tissue under varying compression levels for multiple tissues.
Chapter 6 develops an extended Cole-Cole model to include compression-dependent
terms allowing for an accurate prediction of electrical conductivity of tissue at varying
compressive strains. A strongly non-linear capacitive term is included, indicating the
point at which cell membrane rupture occurs. Moreover, the resistive terms are also
coupled to this capacitive term to properly demonstrate the combining and mixing of
interstitial and intracellular fluids during cell rupture.
Chapter 7 concludes this research and provides ideas for future work in this field
along with the author‟s original contributions.

36. 15
Chapter 2 Thermal Profiling of the Energy-Based Surgical Procedure
2.1 Introduction
The work presented in this chapter has been published previously in The Journal of
Minimally Invasive Gynecology [83].
The use of energized dissection systems has dramatically improved laparoscopic
dissection and hemostasis while allowing more procedures to be performed in a
minimally invasive fashion [84]. Through the years, researchers have worked to improve
this technology in order to enhance its execution both in open and endoscopic cases.
However, thermal collateral damage associated with the use of energized surgical
instrumentation detracts from the usefulness of these devices. The extent of this collateral
damage has been studied previously with varying degrees of cross-correlation [15, 16].
Traditionally the thermal spread from these instruments is determined with a combination
of in-situ dynamic thermography and histopathologic studies [17]. Currently the primary
methods for obtaining hemostasis during minimally invasive surgery include using
ultrasonic energy and radiofrequency current.
Despite reports of minimal thermal spread from today‟s energized instruments, it
(a) (b)
Figure 2.1. (a) Photograph of exteriorized porcine spleen and (b) in situ temperature
acquisition set-up.

37. 16
is impossible for the surgeon to be entirely confident that he or she is not causing
collateral damage. Although thermography in its present form could accompany
energized surgery in open procedures to guide the surgeon, this is not feasible in
laparoscopy. The ability to obtain this data during laparoscopic surgery would require
real-time thermal spread determinations to be made immediately without the use of a
standard thermal imaging camera. If accomplished, this would not only provide the
surgeon with real-time feedback on thermal spread and tissue temperature but also
potentially allow for the cessation of a procedure once critical temperatures have been
reached. Additionally, thermal measurements and thermal spread calculations could be
incorporated into device designs to allow effective control over tissue temperature. The
aim of this preliminary study was to validate the ability of determining sub-surface tissue
thermal profiles in real-time during surgical procedures using either an advanced bipolar
or ultrasonic device.
2.2 Materials and Methods
The setting was an animal surgery operating room at the University of Michigan
School of Medicine. The protocol was approved by the University Committee on Use and
Care of Animals (UCUCA). Funding was secured through an unrestricted educational
grant from Gyrus Medical/ACMI. The experiment was carried out on one large (~50 kg),
white, landrace cross pig.
Anesthesia was induced in the animal with intramuscular injections of telazol (6
mg/kg) and xylazine (2.2 mg/kg), and then the animal was intubated, positioned supine
Figure 2.2. Photograph of top and front views of actual polycarbonate fixtures used for
positioning of both the instrument jaws and the thermistors.

38. 17
on the operating table, and maintained under general anesthesia with isoflurane (2 to
2.5%) while on a ventilator. Oxygen saturation, pulse rate, respiratory rate, mucous
membrane color, and blinking reflex were monitored with pulse oximetry at regular
intervals. Upon completion of the experiment, the animal was euthanized via barbiturate
overdose.
During the course of the operation, a long midline laparotomy incision was made
to expose the abdominal cavity and allow access to the spleen which was the target organ
for the experiment. This decision was based on the target organ‟s size and uniform
thickness. The spleen was exteriorized to perform the instrument measurements and was
replaced within the abdominal cavity during periods of no testing (see Figure 2.1(a)).
Two energized 5 mm laparoscopic devices were used: an ultrasonic instrument as
represented by the Harmonic ACE (Ethicon EndoSurgery, Cincinnati, OH); and an
advanced bipolar electrosurgery instrument as represented by the Gyrus Plasmakinetic
Cutting Forceps (see Table 2.1). The tissue temperature was measured at a depth of 2.0
mm under the tissue surface using thermistors placed at 1.0 mm from each tool edge.
Polycarbonate fixtures were created for each of the devices tested to ensure temperature
measurements were recorded at precise distances from the tool edge (see Figure 2.2 and
Figure 2.3). The additional thermistor ports present in the fixture were incorporated for
future use in acquiring multi-point thermal profiles. Upon tissue clamping the fixture for
each device was placed around the device as shown in Figure 2.1(b) and held in place
while the trial proceeded. Voltage measurements were recorded using a Wheatstone
Bridge circuit and the signals were transmitted and converted in Labview to temperatures
Figure 2.3. Computer Aided Design (CAD) drawing of polycarbonate fixtures
displaying distances of thermistors from instrument jaw edge.

39. 18
via a Data Acquisition System (DAS). A schematic diagram of the complete temperature
data collection system is shown is Figure 2.4.
Modality Ultrasonic Bipolar
Company Ethicon Gyrus
Model Harmonic Ace 5 mm Cutting Forceps
Generator G300 PlasmaKinetic
Setting Min 3 VP-35
Surgical
tool
Top view
of fixture
Front view
of fixture
Fixture
symbol
Within each device, the bite size for the surgical procedures was limited to ¾ of the
jaw length to avoid variations in tissue effect at the jaw hinge area. Lateral tension to the
tissue was avoided and rotational motions were used only when required for by the
device to ensure effects were limited to the devices. Default power settings were used for
each device as listed in Table 2.1.
For the Gyrus Plasmakinetic Cutting Forceps, the tissue was clamped using the built-
in ratchet. Bipolar energy was applied until 4 stars were indicated on the generator
impedance graph. This indicated that a significant change in tissue impedance had
occurred. Further generator details are listed in Appendix A. The seal was then
transected with the Gyrus device‟s cold blade (Figure 2.5(a)). Any trial resulting in less
Table 2.1. Comparison of Two Energized Surgical Instruments.

40. 19
than 3 s to reach 4 impedance stars was omitted as a representative temperature profile
would not have been created in that timeframe.
For the Harmonic ACE, the active jaw was placed under the spleen to allow the cut
process to be applied upwards (Figure 2.5(b)). The tissue was clamped using the ratchet
ensuring it was locked into position. Three clinical scenarios were tested:
Coagulation only – “Min 3” setting applied to only coagulate tissue for a set time
of 5 seconds
Coagulation until cut – “Min 3” setting applied until the surgeon believed
adequate coagulation had occurred and then upward force was applied on the
tissue with the active jaw until the coagulated tissue was divided, still using “Min
3” setting.
Coagulation and then cut – “Min 3” setting to only coagulate tissue for a set
time of 5 seconds and then the “Max 5” setting was used with upward force
applied on the tissue with the active jaw until the coagulated tissue was divided
Figure 2.4. Schematic diagram of the temperature acquisition system used. The signal
conditioner consists of a Wheatstone bridge powered by 10V, with R1=R2=R3=10kΩ.

41. 20
2.3 Results
The Gyrus Plasmakinetic Cutting Forceps was tested in seven separate trials with a
complete set of thermal profiles as shown in Figure 2.6(a). The average maximum
temperature and standard deviation (SD) at 1.0 mm away from the tool edge was 56.8ºC
(SD 8.5ºC) with an average procedural time of 5.7 s (SD 2.7 s).
Three different modes were used with the Harmonic ACE to benchmark its
performance as routinely used in surgical procedures. The device was used to perform a
strict coagulation at the „Min 3‟ setting (Figure 2.6(b)), a coagulation at the „Min 3‟
setting which was continued until a full transection was achieved (Figure 2.6(c)), and
finally a coagulation at the „Min 3‟ setting followed by a transection using the „Max 5‟
setting (Figure 2.6(d)). Each modality was run three times for a total of nine trials. The
average maximum temperatures at 1.0 mm were as follows - coagulation only: 57.5ºC
(SD 7.7ºC); coagulation until cut 61.9ºC (SD 9.6ºC); coagulation and cut 56.4ºC (SD
3.1ºC). The average procedural times were 5.2 seconds (SD 0.9 s) for coagulation only,
10.3 seconds (SD 0.9 s) for coagulation until cut, and 6.0 seconds (SD 0.7 s) for cut.
(a) (b)
Figure 2.5. (a) Post-operative spleen from 5 mm Gyrus Plasmakinetic Cutting Forceps
and (b) post-operative spleen from Ethicon Harmonic Ace.

42. 21
2.4 Discussion
The results in Table 2.2 demonstrate the proof of concept ability to obtain real-time
sub-surface tissue thermal profiles of energized surgical instruments with the novel
temperature acquisition system described. The average maximum temperatures observed
at the 1.0 mm thermistor position for both of the tools tested as well as the average time it
took to achieve them are listed in Table 2.2. Although the Gyrus Plasmakinetic Cutting
Forceps operated cooler than the Ethicon Harmonic ACE at a 1.0 mm distance measured
from the tool edge, the difference in all cases was less than 5ºC and not statistically
(a) (b)
(d)
(c)
Figure 2.6. Thermal Profiles for each of the energized instrument trials performed. (a)
Gyrus 5mm Cutting Forceps, (b) Ethicon Harmonic Ace – „Min 3‟ setting for 5 s, (c)
Ethicon Harmonic Ace – „Min 3‟ setting until tissue transaction, and (d) Ethicon
Harmonic Ace – „Min 3‟ setting for 5 s followed by „Max 5‟ setting until tissue
transaction.

43. 22
Device Application
Tissue Temperature and Surgical Time
Tavg (o
C) TSD (o
C) tavg (s)
tSD
(s)
Gyrus 5mm Forceps
(n=7)
Coag 56.8 8.5 5.7 2.7
Harmonic ACE
(n=9)
Coag (n=3) 57.5 7.7 5.2 0.9
Coag until Cut
(n=3)
61.9 9.6 10.3 0.9
Coag & cut (n=3) 56.4 3.1 6.0 0.7
significant. Also there were no significant differences in temperature amongst the three
ultrasonic energy modes.
The similar thermal range for the two devices would indicate a similar thermal tissue
margin as well; however, the lack of histopathology after the lab precludes this
observation from being made in complete confidence. In fact, previous published reports
have actually shown the level of thermal spread measured via histological analysis to be
less than that of real-time thermography [17]. Based on these reports, collateral thermal
damage may be a function of both the maximum temperature to which the tissue was
exposed and the duration of the application. A more consistent correlation with
histopathology may exist with the use of tissue sub-surface thermistors versus
thermography. Overall, the lack of a difference between the operating temperatures of the
Gyrus Plasmakinetic Cutting Forceps and Ethicon Harmonic ACE demonstrates their
comparable thermal profiles during energized dissection and hemostasis.
2.5 Conclusion
This laboratory experiment represents the first known successful attempt of
measuring sub-surface tissue temperatures in real-time during energized dissection. The
ability to acquire instant temperature information within tissue during an operative
procedure is a major advance not only for the study of surgical tool performance but for
Table 2.2. Comparison of mean peak temperatures, standard deviations, and procedural
times for the 5 mm Gyrus Cutting Forceps and Ethicon Harmonic Ace

44. 23
the safety of surgical procedures in general. Since thermistors are dependent solely on
their own resistance they are believed to be capable of relaying even laparoscopic thermal
information in real-time to the surgeon. This advance has the potential to enhance the
surgeon‟s ability to minimize unwanted thermal damage to surrounding tissues during
surgical procedures if incorporated into instrument designs. Currently, for practical
studies, traditional in situ dynamic thermography with thermal imaging cameras is
limited to open cases.
The initial success achieved in this laboratory experiment warrants the continuation
of this research coupled in the future with histopathologic analysis. The additional
thermistor ports present in the fixture will need to be used to develop multi-point thermal
profiles. Correlating these results with a histopathologic analysis and integration of this
technological concept into instrument design could allow for more accurate determination
of collateral tissue damage during energized dissection and hemostasis thereby
minimizing both intra- and post-operative complications.

45. 24
Chapter 3 Finite Element Model of the Bipolar Electrosurgical
Procedure
3.1 Introduction
The work presented in this Chapter has been published previously in the ASME
Journal of Manufacturing Science and Engineering [85].
In this study, a finite element model (FEM) of the bipolar electrosurgical cautery
procedure is performed to investigate the thermal spread and temperature distribution in
biological tissue. In-vivo surgical experiments are conducted in a porcine model for
temperature measurement in the spleen. The measured temperatures during bipolar
electrosurgery are compared with simulation results to validate the FEM model.
Cautery, or the coagulation of tissue, is a surgical technique that has long been used
to denature proteins and minimize bleeding during surgical procedures [10]. One method
to perform cautery is electrosurgery, which uses radiofrequency (RF) electrical currents
to actively heat biological tissues with high power density [10]. Because human nerve
and muscle stimulation cease at frequencies over 100 kHz, the electrical energy in RF
alternating current can be delivered safely to generate the heat necessary for coagulation.
Without coagulation, the internal bleeding from the cut area is a danger to the patient and
affects the surgeons‟ field of view. Electrosurgery provides a major advance in surgery
by minimizing blood loss and reducing operation time. Miniaturization of the
electrosurgical instrument has enabled the use of minimally invasive or laparoscopic
surgical procedures, thereby reducing patient recovery time. The success of laparoscopic
surgery from both the surgeon and patient perspective has provided the inertia for using
these instruments in increasingly complex procedures. However, their success in
procedures such as prostatectomy and hysterectomy has been hampered by collateral
damage to local neural structures impacting patient post-operative quality of life [3, 4, 7].

46. 25
Electrosurgery can be categorized as monopolar and bipolar. Monopolar
electrosurgery uses current in the gap between tool and tissue to generate heat and ablate
tissue. It functions under the same principle as electrical discharge machining (EDM).
Bipolar electrosurgery employs dispersive electrodes, called forceps, as shown in Figure
3.1. The RF electrical current supplied by an electrical generator flows through only the
tissue between the two electrodes to complete the circuit. As electrical current passes
through tissue its resistance generates heat for cautery. It functions under the same
principles as resistive spot welding. Bipolar electrosurgery is investigated in this study.
Heat generated from electrosurgery has a harmful side effect of spreading and
damaging the surrounding tissue and, more importantly, the nerves in the neurovascular
bundle (NVB) in surgery. This phenomenon is referred to as thermal spread in surgery
[16]. Collateral tissue damage has been highlighted as a major concern for post-operative
side effects, especially for procedures occurring near critical nervous regions such as
prostatectomy and hysterectomy [7]. These side effects include impotence and
incontinence with varying lengths of duration from temporary to permanent. Recent
advances have been made in generator and control technologies that pulse the input
voltage and turn off the power once the tissue has been determined to be coagulated.
Nevertheless, these advances still report thermal spreads of 3–5 mm in ideal situations,
which can cause irreversible side effects during procedures as it is difficult for the
surgeon to control thermal spread from the electrosurgical instrument. The purpose of
this research is to better understand heat transfer in biological tissues in cautery
procedures to further improve the design of surgical bipolar devices.
Thermal spread in biological tissue is difficult to measure and predict. Modeling is a
necessary tool to understand temperature distribution and tissue damage from thermal
spread. However, research is lacking in this area. Research in the modeling of RF
ablation has been reviewed by Berjano [18]. Past research has focused on FEM of tissue
RF ablation and shortening the design time for new RF instrumentation. Several
researchers have modeled RF ablation using a finite element approach [19-21]. However,
the literature has been limited primarily to the area of tumor ablation in the liver and
heart, which is characterized by low voltage inputs and procedural times on the order of
480–720 s and 60-120 s, respectively.

47. 26
Cautery is a technique characterized by high voltage inputs and procedural times on
the order of 3–10 s, almost two orders of magnitude shorter than liver RF ablation and
one order shorter than cardiac RF ablation. To date, detailed FEM on the cautery
procedures is still new and not well studied. Pearce et al. [86] published the finite
difference determinations for the potential gradient from a smooth rectangular electrode.
The modeling of such procedures can be important to further the understanding of how
tissue responds to RF energy, resulting in improvements to instrumentation design. In
this study, a finite element analysis (FEA) of bipolar electrosurgery cautery is performed
to investigate the thermal spread and temperature distribution in biological tissue. In-
vivo surgical experiments are conducted in a porcine model for temperature measurement
in the spleen. The measured temperatures during bipolar electrosurgery are compared
with simulation results to validate the FEM model.
In this study, COMSOL was used to model the heat transfer through in-vivo tissue
during bipolar cautery using the Gyrus ACMI 5 mm Cutting Forceps, as shown in Figure
3.1. The results for temperature-dependent and -independent models are compared to
experimentally measured tissue temperature for validation. In addition, theoretical
compression-dependent effects on electrical conductivity are modeled. The FEM is lastly
used to analyze the thermal profiles of different electrode designs to see how geometry
can be used to reduce thermal spread.
Figure 3.1. The Gyrus ACMI 5 mm bipolar cutting forceps. Note the end of the device
is magnified to show electrode detail.

48. 27
3.2 Experimental Setup for In-Vivo Electrosurgical Temperature Measurement
Experiments were conducted, as shown in Figure 3.2, to measure temperature in the
porcine spleen tissue during an electrosurgical procedure with a Gyrus ACMI 5 mm
Cutting Forceps along with the Gyrus PlasmaKinetic®
generator. As seen from Figure
3.1, the instrument consists of two electrodes, each 13 mm long and 1.15 mm wide. The
distal 20 mm of this probe is made of 301 stainless steel and the proximal 4.0 cm of the
probe is covered with an electrically insulating polytetrafluoroethylene (PTFE) coating.
A porcine (50% duroc, 25% yorkshire and 25% landrace) model weighing about 45
kg was anesthetized and ventilated for use in the in-vivo tissue coagulation experiments.
An Agilent (Santa Clara, CA) 54833A 1 GHz oscilloscope in peak detect mode (PDM)
with an Agilent 10076A 100:1 high voltage probe was used to measure the electrical
voltage input to the tissue. PDM was used due to the limited memory of the oscilloscope
as this only collects the peak readings from each waveform. Data was acquired via the
Labview software. Tissue temperature was measured with micro-thermistors (Alpha
Technics 56A1002-C3, Irvine, CA) with a 0.48 mm outside diameter and 0.25 s thermal
response time. Thermistors were selected over thermocouples and other temperature
sensors due to the high sensitivity and stability in the targeted temperature range
(30−100°C) and their relative immunity to electromagnetic interference. The thermistor
Figure 3.2. Experimental set-up showing positioning of tissue, electrode, and
thermistors.

49. 28
relies on the change in resistance for temperature measurement and is relatively immune
to the significant electromagnetic field generated during electrosurgery.
To maintain the same distance from the cutting edge, a fixture made of
polycarbonate, as shown in Figure 3.2, was custom made that fits the shape of the
forceps. The overall dimension of the fixture is 20 x 38.2 x 5.4 mm. The fixture can
stand firmly on the tissue and has a 2.8 mm tall cavity to allow space for vapor to escape,
a groove in the shape of the surgical tool tip, and several 0.5 mm diameter holes at
specific distances, 1.0, 1.5, 3.0, and 3.5 mm, from the edge of the forceps. The micro-
thermistors were inserted through these holes to measure the temperature inside the tissue
at a set distance and depth in relation to the forceps for comparison to the temperatures
obtained from the FEM.
The electrical input was provided by the Gyrus ACMI PlasmaKinetic®
generator,
commonly used in surgery. Generator details are listed in Appendix A. The measured
AC voltage vs. time between two electrodes in the experiment is shown in Figure 3.3(a).
The voltage signal has two modes. The ON mode is a ±100 V, 350 kHz frequency sine
wave input for about 0.22 s duration. A close-up view of this mode, as illustrated in
Figure 3.3(a), shows the shape and period of the sine waveform. This ON mode is
repeated five times. The second mode is an OFF mode. The voltage signal detected
during this time is a factor of the amplification of noise from the voltage probe, which
uses a 100:1 scaling of the signal in order to protect the oscilloscope from the high
voltage signal.

50. 29
3.3 Finite Element Modeling
3.3.1 Thermal-Electric FEM Formulation
The analytical modeling of heat transfer in tissue, or bio-heat transfer, was pioneered
by the work of Pennes [87] to represent heat sources from metabolism and blood
perfusion. The model was refined and studied extensively in the 70s, 80s, and 90s [88-
91]. The advancement of finite element and finite difference methods in the 90s has
enabled the numerical solution of the bioheat transfer equation for several specific tissue
problems. The bio-heat transfer model of tissue includes coupled thermal and fluid
(blood) transport phenomena. Through the advances in the modeling of bio-heat transfer,
Figure 3.3. Voltage input for FEM (a) measured alternating current (AC) voltage
signal and close up view of the 350 kHz waveform and (b) resultant direct current
(DC) approximation of the waveform equivalent to the root-mean-square (RMS)
value of the RF signal.
0.90
0110
(a) (b)
2.86e-6 s

51. 30
it is assumed that the solid elements can be used to model the tissue, a multi-phase
material consisting of both solid and liquid, with sufficient accuracy.
The linear bio-heat transfer equation for tissue is the general heat equation for
conduction with added terms for heat sources, and can be expressed as [87]:
3-1
where ρ, c, and k are tissue density, heat capacity, and thermal conductivity, respectively,
wb is the effective blood perfusion parameter, cb is the blood heat capacity, T is the local
tissue temperature, Ta is the blood inlet temperature or steady-state temperature of the
tissue, qm is the metabolic heat generation rate of the tissue, qg is the external induced
heat generation rate due to electro-surgical heating of the tissue, and t is time. For all
cases, it was assumed that the metabolic heat source and blood perfusion were
insignificant (qm = 0 and wb = 0) as the energy input into the system is much greater than
that produced during metabolism [92] and the compression of the tissue from the
electrodes inhibits local blood flow. Since the main interest is to simulate the
temperatures achieved throughout a cautery procedure, a time-dependent solution is
considered.
A quasi-static electrical conduction model was applied to solve the electric field in
the tissue using Laplace's equation [93].
3-2
where σ(T) is the temperature-dependent electrical conductivity, and V is the electric
potential. RF coagulation devices operate between 300–550 kHz. At these frequencies,
the wavelength is several orders of magnitude larger than the size of the electrode. Thus,
the majority of the energy generated by the electrosurgical device is dissipated through
electrical conduction rather than capacitive coupling [21].
The difference in the methodology used to solve these governing equations for the
cases of constant and temperature-dependent conductivity stems from the method in
which Laplace's equation (Eq. 3-2) is solved. In the case where the electrical
conductivity is constant, Laplace's equation can be solved independently from the bio-

52. 31
heat transfer equation (Eq. 3-1). The electric potential (V) can be solved quickly over the
entire volume and the solution can be implemented into the source term of the heat
conduction equation. Since temperature varies spatially, temperature-dependent
electrical conductivity is a function of both temperature and position. This dependence
requires that Eqs. 3-1 and 3-2 be solved simultaneously which requires iterative
computation of both the electrical conductivity and temperature.
3.3.2 Properties of Biological Tissue
The electrical and thermal properties of the tissue were available in Refs. [20, 94-96].
The properties for the in-vivo spleen and electrodes are shown Table 3.1. For cases using
a temperature-dependent electrical conductivity, σ(T), a standard increase of 2%/o
C was
used in accordance with Scwhan et al. [97]. Thereby the equations used to determine
σ(T) was:
3-3
where Tref is the baseline temperature for the conductivity.
There has been work performed on the temperature dependent thermal conductivity,
k(T), of porcine spleen by Valvano [96] where the relation can be expressed linearly as:
3-4
Table 3.1. Properties used in the FEM.
Parameter
In-vivo spleen
[20, 94-96]
Electrode [20]
Thermal conductivity ( ref
T
k ) {W/(m·K)} 0.533 70
Density * Specific heat (ρ*c) {J/(K·m3
)} 3.9 x106
2.8 x106
Electrical conductivity ( ref
T
) {S/m} 0.33 4.0x106

53. 32
3.3.3 FEM Techniques
In the multi-physics software COMSOL ver. 3.3, a variation of the heat transfer
equation in the Bioheat Transfer Module enables Eq. 3-1 to be solved. This equation is
coupled simultaneously in the software with the Conductive Media DC Module to solve
Eq. 3-2. The coupling term is the externally induced heat generation term (qg) from Eq.
3-1. This is the resistive heating of the tissue from the RF energy and is defined as qg =
J·E, where J is the current density in the unit of A/m2
and E is the electric field in the
unit of V/m.
Figure 3.4. Schematic of the 3D FEM model showing (a) the tissue, electrodes, and
symmetry plane defined by points EACG , (b) a representative mesh case, (c) top
view of tissue regions identified for the compression-dependent regions (I, II and III)
and thermistor.
(a) (b)
(d)
(c)

54. 33
A schematic of the 3D FEM for this study is shown in Figure 3.4. The grooved
electrode (to be discussed in Section 3.3.4) embedded in the tissue is illustrated in Figure
3.4(a). The original mesh of the tissue was generated using COMSOL‟s automatic
meshing generator and contained 35,269 of the 3D 4-node linear tetrahedral elements.
The mesh refinement feature was used to create a denser mesh in regions near the
electrode where temperature information is critical. The mesh was progressively refined
until the peak temperature solution at 0.5 mm from the side of electrode did not vary by
more than 2%. This resulted in a mesh of 79,476 elements, as shown in Figure 3.4(b),
and all other cases except for the flat electrode (to be discussed in Section 3.3.4) were
performed with this mesh. For each simulation, the electric field (E) and temperature (T)
were calculated.
The COMSOL PARDISO [98] direct solver using matrix row elimination to solve for
the temperature and electrical field was chosen for all analyses. Four input combinations
for electrical conductivity and thermal conductivity material properties were compared in
the model:
1. Constant electrical conductivity and thermal conductivity ( ref
T
and ref
T
k
k ).
2. Temperature-dependent electrical conductivity, constant thermal conductivity
( )
(T and ref
T
k
k ).
3. Constant electrical conductivity, temperature-dependent thermal conductivity
( ref
T
and )
(T
k
k ).
4. Temperature-dependent electrical conductivity and thermal conductivity
( )
(T and )
(T
k
k ).
All variable material properties were determined according to Table 3.1 and Section
3.3.2.
In a separate study, the electrical conductivity of the porcine spleen tissue between
the bipolar electrodes was decreased to simulate the impact of tissue compression.
Preliminary experiments demonstrate that compression influences the tissue electrical
conductivity. This is an area that has limited research but is critical to accurately predict
thermal profiles [99, 100]. Preliminary ex-vivo tests by the authors demonstrate a tissue
electrical resistivity increase of 4 times under 55% compression. In this study, a